> ## Documentation Index > Fetch the complete documentation index at: https://docs.getunbound.ai/llms.txt > Use this file to discover all available pages before exploring further. # Analytics — Overview > Spend, tokens and agent activity at a glance | | | | ----------- | ------------------------------------------- | | **URL** | `/analytics/overview` | | **Sidebar** | Analytics → Overview | | **Roles** | Admin · Manager · Analytics Viewer · Member | Overview summarises your organization's AI usage for the period you select: what it cost, how many tokens it used, and how much your agents did. It is the Analytics tab every role can open. ## What's on the page ### Metric cards Three cards, each with a daily sparkline across the selected range and a comparison against the period before it. | Card | Shows | Underneath | | ------------------------- | ----------------------------------------------------- | ------------------------------------------- | | **Avg Tokens per Prompt** | Average token count across the prompts in this window | The input and output halves of that average | | **Total Estimated Cost** | Estimated spend for the window | Total tokens consumed | | **Agentic Actions** | How many actions your agents took | How many were blocked, and how many warned | ### Charts | Chart | Shows | | --------------------------------- | ------------------------------------------------------------------------------------------------------------- | | **Tokens Spent** / **Cost Spent** | Usage over time, with a **Metric** toggle between **Tokens** and **Cost** | | **Prompts by Model** | The share of prompts going to each model | | **Models, Costs & Tokens** | Cost and token totals for your top eight models, with **Cost** and **Tokens** each toggleable from the legend | | **Agentic Actions by Outcome** | Agent actions over time, split into **Blocked**, **Warned** and **Allowed & Audited** | ### The Analytics tabs The tabs inside the Analytics sidebar section, in the order they appear. | Tab | URL | Admin | Manager | Analytics Viewer | Member | | ------------- | ------------------------ | ------------------------------------- | ------------------------------------- | ------------------------------------- | ------ | | Overview | `/analytics/overview` | | | | — | | Agentic Use | `/analytics/agentic-use` | | | | — | | Data | `/analytics/data` | | — | | — | | Usage · Beta | `/analytics/usage` | | — | | — | | Tasks · Beta | `/analytics/tasks` | | | | — | | Users · gated | `/analytics/users` | | | | — | **Usage** holds two sub-tabs: **Adoption**, which is licensed per organization, and **Account Usage**, which is always available. **Users** appears where your organization has it enabled; elsewhere the tab is absent and its address returns you here. ## Filters and controls ### Time range **24 Hours · 3 days · 7 days · 1 month · 3 months · Custom Range** Remembered between visits, applying to every card and chart. A custom range covers any window within the last **90 days**; once one is applied, the menu's **Custom Range** entry carries a **✓** and the button shows the dates. ### Filters **Add Filter** opens the dimensions available; selections appear as chips and combine. | Filter | Values | | ---------------- | ------------------------------------------------------------------ | | **User** | The people you can read — everyone, for Admin and Analytics Viewer | | **Tool Type** | The AI tools your organization uses | | **Department** | Departments in your organization | | **Team Manager** | Managers in your organization | ## How fresh the numbers are The whole-organization view of a preset range can be up to **30 minutes** behind. Every other view is current: | What you are looking at | Freshness | | ---------------------------------------------------------- | -------------------------------- | | The whole organization, a preset range, no filters | Up to **30 minutes** old | | Any view with a filter chip applied | Live | | Any **Custom Range** | Live | | A Manager's view, which only ever covers their own reports | Live | | A CSV export from an Analytics tab | Current at the moment you export | So the same window can read two ways within a few minutes — an export at 10:00 and the unfiltered page at 10:05 are not measuring the same instant. For live figures, apply a filter or use a custom range covering the same dates. ## What you can do | Action | Result | | ---------------------------- | ---------------------------------------------------------- | | Switch the **Metric** toggle | Changes the series between Tokens and Cost | | Select an Analytics tab | Opens that view with your range and filters carried across | The token series counts input, output and cache-read tokens, while cost prices input, output, cache-read **and** cache-write tokens at each model's rates. That is why the two series on **Tokens Spent / Cost Spent** move independently — and why comparing them is the quickest way to see caching working. ## Messages you may see | Message | What it means | | ------------------------------------------------------------------------------ | --------------------------------------------------------------------------------------- | | `Couldn't load the overview for this window. Try again or adjust the filters.` | The figures for the selected range could not be fetched. Retry, or pick a shorter range | ## Related | Page | For | | ----------------------------------------------------------- | ---------------------------------------- | | [Analytics — Agentic Use](/dashboard/analytics/agentic-use) | What agents did, action by action | | [Analytics — Users](/dashboard/analytics/users) | The same spend broken down per developer | | [Cost Management](/dashboard/cost-management) | Spend against limits, and exemptions |